This article sheds light on research designs, which are strategies researchers employ when conducting scientific research. The research design is chosen in the beginning of a research project, and is determined by the specific issue, problem, or subject under study and the related research questions to explore. Similar to a roadmap, an appropriate research design guides researchers from the outset to the conclusion of a project, without going off tangent or getting lost. The following sections further define the various types of research design, the criteria to consider when selecting the design, and insight into some of the most commonly used research designs.
What is “research design”?
The term research design refers to the framework chosen by a researcher to conduct their research. The design essentially comprises a) the research method(s) best suited to carry out the specific type of research, whether academic or scientific research, and b) the most appropriate techniques to gather the research data. Research designs comprise the complete strategy used to collect, interpret, analyze, and discuss the data obtained to answer research questions and hypotheses.
Given the research design is a framework designed to answer research questions, to determine the best research design for a particular study, researchers need to consider the best data collection method to employ in order to obtain accurate, comprehensive results. This is determined by thinking about the type and primary research goals of the study in question.
There are two main types of research design: qualitative and quantitative. Qualitative methods are employed to collect non-numerical data, such as participant answers to open-ended questions, while quantitative methods collect numerical data and measurable statistics.
While qualitative methods use one-to-one interviews, text analysis, open-ended questions, and case studies, quantitative analysis favors interviews comprising closed question interviews and surveys, statistical data analysis, controlled observations and polls, among others. There are around 19 research types available to researchers, though not all may be applicable. The research problem helps researchers narrow down the most appropriate research type based on the type of data needed.
In a nutshell, a sound research design should fulfill the following criteria:
- State the research problem and any background information as clearly as possible.
- Provide a comprehensive review of the existing literature related to the research topic.
- Clearly state the hypotheses and specific research questions related to the research problem.
- Explain which data is required to answer the research question, and the methodology employed in the data collection process.
- State the techniques to analyze the data. Justify how it is the best method of analysis for the data type, along with its ability to prove the hypotheses.
Elements and characteristics
It is clear that a research design is a comprehensive package containing all the tools a researcher needs to get on with their study, but what exactly does it comprise?
The major elements that make up a research design include:
- The data analysis method employed
- The methodology used in the research study.
- A definitive statement of purpose
- A prediction of the most likely research objections to be raised
- A breakdown of the methods employed to collect and analyze the data
- An outline a proposal timeline, detailing length of study and the key points at each stage of the research.
- A reliable measurement of analysis (i.e., observation and collecting of data through interviews, surveys, etc.)
- A description of the research setting (i.e., physical, social, or experimental setting, such as a neighborhood, a classroom, or a laboratory)
Further, a strong research design should adhere to the following four characteristics:
- Neutrality: To ensure the results are neutral and unbiased, despite any prior assumptions.
- Reliability: Consistency of results is essential, so there should not be any major or unforeseen fluctuations. Well-grounded research questions are key to obtaining satisfactory results.
- Validity: Precise and appropriate tools to measure the data are intrinsic to the validity of the study and ensuring comprehensive data analysis.
- Generalization: Even if the research is conducted on a small study population, the results should still apply to the general population with equal accuracy.
Types of research designs
There are many different research designs available to researchers. While the majority fall into the qualitative and quantitative categories, they can be further categorized into five types:
A descriptive research design is the most general type of research. It principally involves observing the collection, analysis, and interpretation and presentation of data. As the researcher observes situations as they are, potential variables are disregarded. Descriptive research designs are used to determine change, opinions and preferences, similarities and differences, and quantities.
- Correlational: The aim of a correlational research design is to identify correlations between variables and non-causal relationships and test the reliability of measurement tools. The standard correlations are positive, negative, and zero, which allows for faster data collection times.
- Experimental: Experimental research designs are typically used to determine cause and effect. The experiment can be conducted in the field, in a controlled environment, or be a quasi-experiment. Researchers generally obtain the research data by measuring the changes to dependent variables caused by alterations in independent variables.
- Diagnostic: Diagnostic research designs aim to find out why something happens. Diagnostic research is undertaken to find the potential causes to problems. Hence, researchers follow a three-stage research plan to find out the inception, diagnosis, and solution of a problem.
- Explanatory: Researchers use explanatory research designs to investigate, expand, and explain an existing problem or specific area of research. It is investigatory by nature, with the primary aim to clarify little-understood theories or ideas.
Let’s take a further look at some other common research design types to get an understanding of how the various designs, groups, and sub-groups fit into specific categories.
Action Research Design
As the saying goes, there’s no change without action. This type of research is focused on precisely that-change-which takes place through participation, often with groups of participants.
The researcher will often distribute questionnaires, interview questions, or surveys to not only determine individual opinions and responses, but they will also reflect on the group’s responses as a while. Action research is typically applied in organizations and in academic research settings, where researchers can observe and reflect on processes and responses in real-time.
Case studies are in-depth investigations into theories and models to determine their validity in real-world scenarios. The researcher may use several research methods and sources to conduct analysis from various angles.
The case study research design is often employed to explore under-explored research areas. Common fields in which researchers opt for case study research designs include the education, social work, medicine, anthropology, and psychology.
Causal designs focus on cause and effect. This design is explanatory in nature, with the aim to find out the extent of the relationship between the variables.
Causal designs are useful when seeking to identify reasons for specific outcomes and what changes have come about following an event. They generally allow for accurate results, though coincidences may also occur, and can be hard to detect. A few common subjects investigated in causal studies include personality, preferences, spending habits, behavior, performance in business and academia, etc.